Methodology for the objective non-invasive diagnosis of functional and anatomical brain lesions will be developed. Brain electrical potentials from multiple scalp electrodes will be recorded from patients with supratenotorial brain tumors and from normals. Spectral analysis will be performed on resting electroencephalographic activity and sensory evoked potentials will be formed. Topographic maps of the spectral and evoked potential data will be created. An interactive development and evaluation of features descriptive of the topographic, spectral, and dynamic course of brain electrical potentials will be followed by automated information extraction and computer processing of features by self-learning computer algorithms. Sensitive measures of symmetry of feature distributions will be employed. Objective characterizations of the locations and size of brain tumors (CAT scan) will be derived from the topographic features. Objective characterizations of the location and magnitude of the associated functional deficits (neuropsychological testing) will also be derived. A data base of topographic images will be collected and a subcategorization of different diagnostic categories, based on objective criteria, will be established. This automated diagnostic methodology will be compared to more conventional diagnostic techniques.